Here are several ways to optimize training performance:

• Add more images per class : ≥ 1500 images per class recommended.
• More image variety : recommend images from different times of day, different seasons, different weather, different lighting, different angles....
• Select larger models : like yolov8l, yolov8x....
• Increase Epoches : Start with 300 epochs. If overfitting does not occur after 300 epochs, train longer, i.e. 600, 1200 etc. epochs.
• Larger image size : the default training image size (imgsz) is 640, magnify the image size for more accuracy.

What I’ve shared are just a few of the available solutions. The YOLO website offers more comprehensive and detailed strategies for improving training performance. You can refer to the following link: https://github.com/ultralytics/ultralytics/blob/main/docs/en/yolov5/tutorials/tips_for_best_training_results.md